Evaluating Localization Algorithms for Outflow Tract Premature
Ventricular Complexes in Patients Performed Ablation Therapy- Insights
from a Single-Center Study
Abstract
Introduction: Numerous algorithms have been published to
estimate the origin of Premature Ventricular Complexes PVCs from the
outflow tract, and many studies have compared them. However, a clear
consensus on the best algorithm has not yet been established. We aim to
assess and compare algorithms for locating the origin of Premature
Ventricular Complexes (PVCs) in ECG recordings from patients undergoing
outflow tract PVC ablation. Methods and Results: We analyzed
ECG records from 116 patients who underwent successful PVC ablation from
the outflow tract between June 1st, 2015, and June 30th, 2020. Of these
patients, 53 had PVCs originating from the right ventricular outflow
tract (RVOT), and 63 had PVCs originating from the left ventricular
outflow tract (LVOT). R wave amplitude index in V1 combined with R wave
deflection interval in V3, V2 transition ratio, V2S/V3R index,
Transition Zone (TZ) İndex, Combined TZ and V2S/V3R index, S and R wave
difference in V1 and V2 and R wave duration index and R/S wave amplitude
index algorithms and their components were analyzed. The Combined TZ
Index algorithm had the maximum sensitivity and negative predictive
value, which were 90.48% and 85.37%, respectively. The R Wave Duration
Index and R/S Amplitude Index algorithms had the highest specificity and
positive predictive value, which were assessed at 98.11% and 94.74%,
respectively. With the logistic regression analysis method, the
“Y=8.436 – 2.036 x PVC TZ score- 0.06 x PVC V2 R wave duration + 4.661
x PVC V3 R wave amplitude - 1.958 x PVC V2 S wave amplitude” algorithm
was created. In this algorithm, the value of the “Y” variable was
accepted as >0.5 for LVOT. Conclusion: The Combined
TZ Index algorithm had the maximum sensitivity and negative predictive
value and the R Wave Duration Index and R/S Amplitude Index algorithms
had the highest specificity and positive predictive value in our
population (85.71%, 85.37%, 92.45%, 89.47% respectively). “Y=8.436
– 2.036 x PVC TZ score- 0.06 x PVC V2 R wave duration + 4.661 x PVC V3
R wave amplitude - 1.958 x PVC V2 S wave amplitude” algorithm predicts
LVOT if Y>0.5.